NHE DEFLATORINTERMEDIATE SUMMARY

National Health Expenditures

National Health Expenditures (NHE) in the United States include all spending related to the purchase of health care
1
goods and services during the year and the amount invested to procure future health services. Historically, U.S.
health care spending has grown faster than most other sectors of the economy. Differences in the annual growth
rates of the NHE reflect trends in the factors that drive health care spending, including:
1.
2.
3.
4.

Increases in technological developments

Changes in the age and sex composition of the population (demographic effect)
Changes in the use and mix (or intensity) of health care services
Changes in prices for health care goods and services

The NHE has traditionally been reported in nominal terms (current dollar) and has not been adjusted to remove the
impact of changes in health care prices (constant or real dollars). Although a price index for Personal Health Care
(PHC) goods and services has been available for many years, there has not been, until now, a corresponding price
index for the aggregate NHE. The Office of the Actuary in the Centers for Medicare and Medicaid Services (CMS)
released its new chain-weighted NHE price deflator with the publication of the 2011 NHE Accounts (NHEA). The new
chain-weighted NHE price deflator, which is available for 2004 2011, allows for the analysis of total health
spending in real terms. Real estimates of health sending can now be compared to trends in underlying non-price
factors such as population, utilization, intensity, mix of goods and services, and demographics. This document
briefly describes the method used to develop the chain-weighted NHE price deflator.
NHE Deflator Calculation
The NHE price deflator is a chain-weighted index that uses a wide range of detailed price indexes from the Bureau of
Labor Statistics (BLS) Consumer Price Index (CPI) and Producer Price Index (PPI) programs. The NHE Deflator is an
aggregate price index; therefore, its critical that it not capture the biases that can occur when aggregating individual
price indexes. The chain-weighted method used in the NHE Deflator attempts to control for any aggregation bias by
using a Fisher Ideal formulation. The Fisher Ideal index formulation reflects the geometric mean of a Laspeyres
index, which uses prior period quantity weights, and a Paasche index, which uses current period quantity weights. As
a result, chain-weighted price measures typically yield lower inflation rates than standard indexes (such as Laspeyres
or Paasche) since substitutions are made over time to purchase less of the goods or services that experience faster
price growth. Equation (1) below is the formulation of the NHE Deflator using a Fisher Ideal index, where the first
term represents the Laspeyres price index change and the second term represents the Paasche rice index change:
(1)

) (

= time period
= NHE categories
= NHE deflator
= Price index for category
= Quantity Index for category
1

For further information, see National Health Expenditure Accounts: Methodology Paper, 2011 [Definitions, Sources, and Methods]
Page 6 What are the National Health Expenditures? http://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-andReports/NationalHealthExpendData/downloads/dsm-11.pdf

Chaining together the period-by-period changes in the NHE Deflator (i.e. the Fisher ideal formulation) ensures that
compositional changes in the quantity weights are kept distinct from the price changes. Under the chain-weight
method this occurs because the base period quantity weights for the Laspeyres formulation is always only one
period prior to the base period quantity weights for the Paasche formulation. The chain-weighted percent change
for period t+1 is determined as
, the percent change for period t+2 is determined as
, and so on. This chain-weighted method is the preferred index formulation used by the
Bureau of Economic Analysis (BEA) when deflating the National Income and Product Accounts.
NHE Components
The NHE can be divided into two major components: PHC and Non- PHC. PHC measures the total amount spent to
treat individuals with specific medical conditions and includes 10 categories of goods and services (Table 1) such as
hospital care, physician and clinical services, and retail prescription drugs. Non-PHC includes government
administration, the net cost of private health insurance, government public health activity, investment in research,
and investment in structures and equipment. The components of PHC can be deflated using specific price indexes
from the BLS that are associated with the medical goods and services provided. However, estimation of prices for
the Non-PHC components of the NHE are more complicated because there may not be available price indexes for
these types of health spending as they typically dont involve market transactions made by an individual or on an
individuals behalf.
Table 1: NHE, PHC and Non-PHC categories
Sub-aggregate
Personal Health Care
Hospital Care
Physician & Clinical
Other Professional Services
Dental Services
Other Health, Residential, and Personal Care Services
Home Health Care
Nursing Home Care
Prescription Drugs
Other Non-Durable
Durable Medical Products
Non-Personal Health Care
Government Administration
Net Cost of Insurance
Government Public Health
Research
Structures and Equipment

Price Series
Availability

Matches
NHE Concept

Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes

Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes

No
Yes
No
Yes
Yes

No
No
No
Yes
Yes

In instances where price indexes arent readily available (or the indexes do not match the NHE concept) for the NonPHC components of NHE, we have constructed composite input price indexes that match the spending categories
within each Non-PHC component. These new composite indexes, described in detail later, are used to deflate each
Non-PHC component.
Personal Health Care
The PHC deflator is calculated as a chain-weighted price index for the various goods and services that account for
PHC spending. Table 2 lists the detailed price series that are used for each component of PHC expenditures.

Table 2: PHC Spending Components, Nominal Share of NHE, and Associated Price Proxies
Sub-aggregate
Hospital Care
Physician & Clinical

Nominal Share
of 2011 NHE
31%
20%

Other Professional Services

Dental Services
Other Health, Residential,
and Personal Care Services

3%
4%
5%

Home Health Care

Nursing Home Care
Prescription Drugs
Other Non-Durable
Durable Medical Products

3%
6%
10%
2%
1%

Total

84%

Price Series
PPI, hospitals
Composite Index:
PPI, offices of physicians
PPI, medical and diagnostic laboratories
CPI, services by other medical professionals
CPI, dental services
Composite Index:
CPI physician services
CPI care of invalids and elderly at home
CPI All Items
PPI residential mental retardation facilities
PPI, home health care services
PPI, nursing care facilities
CPI, prescription drugs
CPI, non-prescription drugs
Composite Index:
CPI, eyeglasses and eye care
CPI, medical equipment and supplies

CY2011 weights are shown in Table 2, but in formulating the NHE Deflator, weights are varied by year as part of the
chain weight calculation.
Table 3 shows the average annual growth in nominal and real PHC spending and growth in the chain-weighted PHC
price index.
Table 3: PHCAverage Annual Percent Change from Preceding Year shown for Nominal PHC Spending, Aggregate
PHC Price Index, and Real PHC Spending
Item
PHCNominal
PHCChain Weighted Price Index
PHCReal

1990
11.0
7.1
3.7

2000
6.6
3.0
3.5

2003
8.3
3.2
4.9

2004
7.4
3.5
3.7

2005
6.8
3.1
3.6

2006
6.3
3.1
3.2

2007
6.1
3.4
2.6

2008
5.0
2.6
2.4

2009
5.0
2.8
2.2

2010
3.7
2.7
1.0

2011 1960-2011
4.1
9.4
2.1
4.9
2.0
4.2

In Table 3, 1990 data reflects average annual growth from 1980 to 1990.
Non-Personal Health Care
Unlike the PHC deflator, where one price series is normally used to represent the pure price change associated with
the entire category, the non-PHC categories are typically deflated by an input price index that represents the price
increases associated with the expenses underlying the production of these categories (the notable exceptions are
non-commercial research and structures and equipment). Because of the unique nature of the non-PHC categories,
there are typically not publicly available price series for these categories, or those that are available do not
adequately capture the concepts appropriate for the given non-PHC category. Instead, alternative data sources are
used to decompose these categories into the key underlying inputs used in their production, such as compensation
or capital costs, and then publicly available price series are used to deflate those input costs. A brief description of
each price deflator follows.

Non PHCGovernment Administration

Government administrative spending is deflated using a composite input price index that chain-weights together
price indexes for wages and salaries, benefits, professional fees, claims processing services, office rent and other
expenses. The input weights reflect six sub-categories of government administrative spending: Medicare,
Department of Defense (DOD), Veteran Affairs (VA), Medicaid, Childrens Health Insurance Program (CHIP) and other
third party payers (OTP).
The government administrative input price index is composed of two sub-aggregate indexes: a federal input cost
index (Medicare, DOD and VA costs) and a state and local input cost index (Medicaid, CHIP and OTP costs). The cost
weights for these indexes are determined using data from the Medicare Trustees Report, Medicaid administrative
data, and Congressional budget justifications for CMS and SSA. The price series (Table 4) used for each of the
categories represent proxies for their respective concepts, such as federal civilian pay, Employment Cost Indexes
(ECIs) for state and local government workers, ECIs for other relevant occupations, PPIs and CPIs.
Cost weights for the federal index are calculated based on Medicare data; we assume that the DOD and VA costs
reflect a similar distribution as that calculated for Medicare. Cost weights for state and local administration are
calculated based on Medicaid data; we assume that the CHIP and OTP costs reflect a similar distribution as that
calculated for Medicaid.
Table 4: Government AdministrationComponents, Nominal Share of 2011 NHE, and Associated Price Proxies
Sub-aggregate
Price Series
Federal Costs (0.4% nominal share of 2011 NHE)
Wages
Federal Civil Service Pay Adjustments (Congressional Research Service)
Benefits
ECI for Total Benefits, All Workers, Private Industry (BLS)
Security, IT, Training, and Other
ECI for Total Compensation, Professional, Scientific and Technical
Services
Services (BLS)
Claims Processing and Financial
ECI for Total Compensation, Insurance Carriers, and Related Activities
Intermediaries
(BLS)
Rent
PPI for Leasing of Professional and Office Buildings (BLS)
Other
CPI-Urban for All Items (BLS)
State and Local Costs (0.8% nominal share of 2011 NHE)
Compensation
ECI for Total Compensation, State and Local Government, Public
Administration (BLS)
IT and Other Services
ECI for Total Compensation, Professional, Scientific and Technical
Services (BLS)
Claims Processing and Eligibility
ECI for Total Compensation, Insurance Carriers, and Related Activities
Determination
(BLS)
Other
CPI-Urban for All Items (BLS)

CY2011 weights are shown in Table 4, but in formulating the NHE Deflator, weights are varied by year as part of the
chain weight calculation.
Table 5 Government AdministrationAverage Annual Percent Change in Nominal Spending from Preceding Year
shown, Aggregate Price Index, and Real Spending
Item
Government AdministrationNominal
Government AdministrationChain
Weighted Price Index
Government AdministrationReal

1990
10.0

2000
9.1

2003
12.8

2004
7.6

2005
7.4

2006
1.5

2007
2.4

2008
2.9

2009
1.8

2010
0.7

2011
4.7

3.6

3.4

3.1

3.3

3.6

1.1

1.8

2.4

3.9

3.8

-1.5

-0.9

-0.6

0.7

-1.1

2.2

In Table 5, 1990 data reflects average annual growth from 1980 to 1990.

Non PHCNet Cost of Health Insurance

The net cost of health insurance input price deflator is a chain-weighted composite index of input costs and price
proxies designed to measure the price growth associated with the net cost of insurance, which is the difference
between health insurance premiums earned and benefits incurred. This difference includes costs such as
administrative services, taxes, and changes to reserves and underwriting gains or losses. The types of private health
insurance for which net cost is estimated include: fully-insured group/commercial, individually-purchased or nongroup, self-insured group, and the health portion of property and casualty. We also included the net cost from the
following types of insurance: Medicare Advantage and stand-alone Medicare Part D plans; Medicaid managed care
plans; CHIP managed care plans; the majority of workers compensation insurance; and, in 2010, the pre-existing
condition insurance plan.
Cost categories for net cost of health insurance are determined using data primarily from AM Best, which is based on
National Association of Insurance Commissioners (NAIC) insurance statements. These cost categories are:
compensation of the employees that are administering the insurance; capital costs; taxes; other costs (such as rent,
advertising, certain commissions, etc.); and, in some cases, changes to reserves and underwriting gains or losses. We
developed an input price index for the net cost of health insurance using these five cost categories. A blended index
of price proxiestypically ECIs, PPIs, or in some cases price indexes from the Gross Domestic Product accountsare
weighted together by the respective input costs for three of these general cost components (compensation, capital,
and other costs). All changes in taxes, reserves, or underwriting gains or losses are treated as price changes. We then
combined these price changes to create a composite net cost of health insurance input price deflator.
Table 6: Net Cost of Health InsuranceComponents and Associated Price Proxies
(5.8% nominal share of 2011 NHE)
Sub-aggregate

Price Series
ECI for Total Compensation, Insurance and Related Activities
Composite Index:
ECI for Total Compensation, Insurance and Related Activities
PPI, Commissions from Insurance
PPI, Legal Services
PPI, Advertising
CPI, Postage
GDP, Implicit Price Deflator
Price Index, Non-Residential Equipment & Software (BEA Table 5.5.4)
None (no change to price in real terms)
None (no change to price in real terms)

Compensation
Other Expenses

Capital Related Expenses

Taxes
Changes in Reserves and Underwriting Gains or Losses

CY2011 weights are shown in Table 6. In formulating the NHE Deflator, weights are varied by year as part of the
chain weight calculation.
Table 7: Net Cost of Health InsuranceAverage Annual Percent Change in Nominal Spending from Preceding Year
shown, Aggregate Price Index, and Real Spending
Item
Net Cost of Health Insurance
Nominal
Net Cost of Health Insurance
Chain Weighted Price Index
Net Cost of Health Insurance
Real

1990

2000

2003

2004

2005

2006

2007

2008

2009

2010

2011

13.1

7.3

18.7

6.9

6.9

11.7

4.0

-2.0

-1.7

9.8

4.0

2.7

7.7

4.8

3.6

-6.8

1.5

7.4

7.2

4.1

-0.7

6.5

0.4

5.1

-3.2

2.2

-3.0

In Table 7, 1990 data reflects average annual growth from 1980 to 1990.

Non PHCGovernment Public Health Activities

Public Health spending in NHEA is deflated using a composite index that chain-weights together price indexes for
state and local and federal public health, with state and local expenditures accounting for roughly 80 percent of the
index. State and local public health expenditures are deflated using the price index for gross state and local
government consumption expenditures for health from the National Income and Product Accounts produced by the
BEA. Federal public health expenditures are deflated using an input price index that weights together the input costs
of Health Resources and Services Administration, Food and Drug Administration, and Centers for Disease Control
and Prevention; and appropriate price proxies from the BLS. Together these three organizations account for over
75% of federal public health spending.
Table 8: Government Public Health ActivitiesComponents and Associated Price Proxies
(2.9% nominal share of 2011 NHE)
Sub-aggregate

Price Series

Federal
Wages
Benefits
Security, IT, Training, and Other Services
Rent
Other
State and Local
[none]

Federal Civil Service Pay Adjustments (Congressional Research Service)

ECI for Total Benefits, All Workers, Private Industry (BLS)
ECI for Total Compensation, Professional, Scientific and Technical Services (BLS)
PPI for Leasing of Professional and Office Buildings (BLS)
CPI-Urban for All Items (BLS)
BEA Price Index for Gross State and Local Government Consumption Expenditures

CY2011 weights are shown in Table 8, but in formulating the NHE Deflator, weights are varied by year as part of the
chain weight calculation
Table 9: Government Public Health ActivitiesAverage Annual Percent Change in Nominal Spending from
Preceding Year shown, Aggregate Price Index, and Real Spending
Item
Government Public Health Activities
Nominal
Government Public Health Activities
Chain Weighted Price Index
Government Public Health Activities
Real

1990

2000

2003

2004

2005

2006

2007

2008

2009

2010

2011

12.0

8.0

7.5

0.5

4.1

11.2

10.3

5.8

4.1

4.9

-0.5

4.0

4.8

4.2

4.1

5.1

0.4

2.5

3.4

-3.4

-0.7

6.8

5.9

0.6

3.7

2.4

-3.7

In Table 9, 1990 data reflects average annual growth from 1980 to 1990.
Non PHCNon Commercial Research
We deflate non-commercial research using the Biomedical Research and Development Price Index (BRDPI), which is
developed and updated annually by the BEA under an interagency agreement with the National Institutes of Health
(NIH). The BRDPI is designed to measure changes in the weighted-average of the price of all the inputs (wages,
equipment, nondurables, etc.) purchased with the NIH budget in support of extramural research. Over two-thirds of
non-commercial research in the NHEA is conducted by NIH, making this index a reasonable choice for the deflation
of non-commercial research. There are no sub-aggregate cost categories for non-commercial research.
In order to deflate non-commercial research spending, the nominal level of spending is adjusted by the BRDPI index
to produce real non-commercial research spending.

Table 10: Non Commercial ResearchComponents, Nominal Share of 2011 NHE, and Associated Price Proxies
Item
Non-Commercial Research

Nominal Share of
2011 NHE
1.8%

Price Series
Price Index, NIH Biomedical Research and Development

CY2011 weights are shown in Table 10, but in formulating the NHE Deflator, weights are varied by year as part of the
chain weight calculation
Table 11: Non Commercial ResearchAverage Annual Percent Change in Nominal Spending from Preceding Year
shown, Aggregate Price Index, and Real Spending
Item
Non-Commercial Research
Nominal
Non-Commercial Research
Chain Weighted Price Index
Non-Commercial Research
Real

1990

2000

2003

2004

2005

2006

2007

2008

2009

2010

2011

8.9

7.2

11.0

10.5

4.7

2.6

1.3

3.5

4.3

8.2

1.7

3.7

3.9

4.4

4.0

4.5

3.4

2.9

2.8

6.5

0.8

-1.7

-2.6

-1.0

1.0

5.2

-1.1

In Table 11, 1990 data reflects average annual growth from 1980 to 1990.
Non PHCStructures & Equipment
Investment in structures and equipment is deflated using a composite index that chain-weights together price
indexes associated with private fixed investment in structures and equipment by asset category. The nominal
investment levels by asset category serve as the component weights. These detailed investment levels are obtained
by decomposing overall private and public nominal investment in structures and equipment in the NHEA using
primarily the BEAs Capital Flow Table (CFT) and Fixed Asset Accounts (FAA).
Five categories of investment in Structures and twenty-two categories of investment in Equipment are derived. The
largest categories can be seen below in Table 13. Price indexes are then selected for each of these Structures and
Equipment categories. The price indexes for private investment are from BEA Table 5.4.4. Price Indexes for Private
Fixed Investment in Structures by Type and from BEA Table 5.5.4. Price Indexes for Private Fixed Investment in
Equipment and Software by Type.
Table 12: Structures and EquipmentComponents, Nominal Share of 2011 NHE, and Associated Price Proxies
(3.8% nominal share of 2011 NHE)
Sub-aggregate
Structures (1.7% share)
Hospital and institutional buildings
Other
Equipment (2.2% share)
Computers and Peripheral Equipment
Total Software
Medical Equipment and Instruments
Light trucks (including utility vehicles)
Other

Price Series
BEA Price Index, Table 5.4.4 Line 5
BEA Price Index, Table 5.4.4 Lines 4, 14, 17, and 23
BEA Price Index, Table 5.5.4 Line 5
BEA Price Index, Table 5.5.4 Line 6
BEA Price Index, Table 5.5.4 Line 8
BEA Price Index, Table 5.5.4 Line 21
BEA Price Index, Table 5.4.4 Lines 7, 9, 10, 11, 13-18, 22-24, 28, 30, 32-34

CY2011 weights are shown in Table 12, but in formulating the NHE Deflator, weights are varied by year as part of the
chain weight calculation.

Table 13: Structures and EquipmentAverage Annual Percent Change in Nominal Spending from Preceding Year
shown, Aggregate Price Index, and Real Spending
Item
Structures and Equipment
Nominal
Structures and Equipment
Chain Weighted Price Index
Structures and Equipment
Real

1990

2000

2003

2004

2005

2006

2007

2008

2009

2010

2011

9.4

5.6

6.6

5.1

9.2

3.9

13.6

8.6

-8.7

-0.7

3.6

2.0

3.0

3.4

2.0

1.1

-1.6

-3.0

0.7

3.0

6.0

0.5

11.3

7.4

-7.2

2.4

2.9

In Table 13, 1990 data reflects average annual growth from 1980 to 1990.
NHE Deflator
As previously described, the PHC and Non-PHC components are chain-weighted together in order to calculate the
NHE Deflator. Table 14 shows the final chain-weighted NHE Deflator, as well as nominal NHE and the resulting real
NHE. Figure 1 shows the same data graphically from 2004 to 2011.
Table 14: NHE DeflatorAverage Annual Percent Change in Nominal Spending from Preceding Year shown,
Aggregate Price Index, and Real Spending
Item
NHENominal
NHE DeflatorChain
Weighted Price Index
NHEReal

1990
11.0

2000
6.6

2003
8.8

2004
7.1

2005
6.8

2006
6.5

2007
6.2

2008
4.7

2009
3.9

2010
3.9

2011
3.9

3.4

3.4

3.2

3.3

2.1

2.4

2.7

2.4

3.5

3.2

3.2

2.8

2.6

1.5

1.2

1.5

In Table 14, 1990 data reflects average annual growth from 1980 to 1990.
Figure 1: Annual Percent Change in Nominal NHE, NHE Price Index, and Real NHE, 2004 to 2011
8.0
Nominal
Price Index (chain weighted)
Real

Annual Percent Change

7.0
6.0
5.0

4.0
3.0
2.0
1.0

0.0
2004

2005

2006

2007

2008

2009

2010

2011

As Figure 1 indicates, health care price growth as measured by the NHE Price index accounts for slightly more than
half of the growth in nominal health spending from 2004 to 2008. From 2008 to 2011, health care price growth
accounts for approximately two-thirds of growth in nominal health care spending. The contribution of prices to NHE
growth fluctuates by year, from a low of about 45% in 2008 to a high of almost 70% in 2010. The variations in these
trends are heavily influenced by the Non-PHC prices that underlie the NHE deflator, as shown in Figure 2.

Figure 2: Annual Percent Change in the NHE Price Index, the PHC Price Index, and the Non-PHC Price Index, 2004
to 2011
6.0
NHE Deflator
PHC Deflator
Non-PHC Deflator

Annual Percent Change

5.0
4.0
3.0
2.0
1.0
0.0

-1.0
2004

2005

2006

2007

2008

2009

2010

2011

Growth in the non-PHC deflator outpaced growth in the PHC deflator during the 2005-2006 period, in large part due
to the rapid price growth for the net cost of insurance, government public health, and research. However, this trend
reversed in 2008 and 2009. Prices for non-PHC decreased in 2008, mainly the result of declines in the net cost of
insurance as the overall economic recession impacted underwriting gains, losses, and reserves. In 2009, the cost of
government administration; government public health, structures and equipment; and the net cost of insurance
grew slowly or decreased. In 2011, price growth for non-PHC was higher than PHC, mainly due to increased price
growth of net cost of insurance due to the economic recovery.